How we move is universal: scaling in the average shape of human activity

TL;DR

This study reveals a universal scaling law in human motor activity fluctuations across various temporal scales, based on a month of wristwatch data, with potential applications in health assessment.

Contribution

It demonstrates the existence of a universal law governing human activity fluctuations across multiple time scales, extending concepts from physical systems to human behavior.

Findings

Identified universal critical exponents in activity fluctuations

Observed similarity to crackling noise phenomena

Potential for new health biomarkers

Abstract

Human motor activity is constrained by the rhythmicity of the 24 hours circadian cycle, including the usual 12-15 hours sleep-wake cycle. However, activity fluctuations also appear over a wide range of temporal scales, from days to a few seconds, resulting from the concatenation of a myriad of individual smaller motor events. Furthermore, individuals present different propensity to wakefulness and thus to motor activity throughout the circadian cycle. Are activity fluctuations across temporal scales intrinsically different, or is there a universal description encompassing them? Is this description also universal across individuals, considering the aforementioned variability? Here we establish the presence of universality in motor activity fluctuations based on the empirical study of a month of continuous wristwatch accelerometer recordings. We study the scaling of average fluctuations across temporal scales and determine a universal law characterized by critical exponents $\alpha$, $\tau$ and $1/{\mu}$. Results are highly reminiscent of the universality described for the average shape of avalanches in systems exhibiting crackling noise. Beyond its theoretical relevance, the present results can be important for developing objective markers of healthy as well as pathological human motor behavior.